I have Windows 7, and I have a GTX 580 used only for computation and an 8800 used only for display. The PhysX acceleration is configured to use the 8800 card.
I am using CUDA 4.2 with the latest driver (301.32). I am still getting timeout (TDR) errors. These errors go away if I edit the registry, but we would like to not require our customers to hack the registry, and we’d also like to not break our kernels into sub two-second intervals.
I had been under the impression that non-display cards were not subject to TDR problems. Am I mistaken? Thanks for any help.
Update: Found this in the release notes:
- (Windows and Linux) Individual GPU program launches are limited to a run
time of less than 5 seconds on a GPU with a display attached. Exceeding this
time limit usually causes a launch failure reported through the CUDA driver
or the CUDA runtime. GPUs without a display attached are not subject to the
5 second runtime restriction. For this reason it is recommended that CUDA
be run on a GPU that is NOT attached to a display and does not have the
Windows desktop extended onto it. In this case, the system must contain at
least one NVIDIA GPU that serves as the primary graphics adapter.